Background/Aims
Tropical teleost species in northern Australia typically exhibit rapid growth with asymptotic length reached at young ages. Northern Australian fisheries often implement strong length selectivity through gear selection and/or legal minimum lengths (LML), particularly in multi-sector fisheries which impacts the biological data available for population assessment. We present a case study for black jewfish (Protonibea diacanthus) from the Northern Territory where independent growth estimation differed significantly from growth estimated within a stock assessment model that simultaneously estimated length and age-based selectivity.
Methods
Black jewfish samples were collected from the Greater Darwin region during 2005-2007 and 2019-2021. A Bayesian von Bertalanffy growth function was fit to these samples using the 'BayesGrowth' R package. Growth was also estimated within a statistical catch-at-age model applied through Stock Synthesis, using BayesGrowth parameters as starting values with an additional prior for L∞.
Results
Samples from 2005-2007 and 2019-2021 showed strong length selectivity, with most specimens between 90-125 cm TL, though some smaller fish (35-90 cm TL) were included. The sampling gear matched that of the NT Coastal Line Fishery (CLF) and is considered representative of this fishery. The BayesGrowth model fitted strongly to modal lengths for each age class despite significant variation in length-at-age for ages one to three, producing the typical fast growth trajectory of tropical species. Conversely, Stock Synthesis produced a slower growth trajectory with an asymptote at older ages. The selectivity estimates confirmed that sampling was highly length-selective, with slower-growing fish underrepresented in the length-at-age data for two-to-three-year age classes. However, the Stock Synthesis growth curve accounted for these under sampled fish in the estimated growth parameters.
Conclusion
Length selectivity restricted samples for one-to-three-year-olds to the fastest-growing individuals, producing a growth curve with faster k and an L∞ that was reached at a younger age. Simultaneous estimation of length and age selectivity through Stock Synthesis overcame this sampling limitation, accounting for under sampled slower-growing fish at younger ages. This demonstrates the impact of length selectivity on growth estimation and the value of accounting for it through stock assessment modelling or other methods.